AUTHOR=Su Qiaomei , Tao Weiheng , Mei Shiguang , Zhang Xiaoyuan , Li Kaixin , Su Xiaoye , Guo Jianli , Yang Yonggang 

TITLE=Landslide Susceptibility Zoning Using C5.0 Decision Tree, Random Forest, Support Vector Machine and Comparison of Their Performance in a Coal Mine Area

JOURNAL=Frontiers in Earth Science

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2021.781472

DOI=10.3389/feart.2021.781472

ISSN=2296-6463

ABSTRACT=The main purpose of this study is to establish an effective landslide susceptibility zoning model and test whether underground mined area and surface collapses in coal mine areas affect seriously on the occurrence of landslides. Taking Fenxi Coal Mine Area of Shanxi Province in China as the research area, landslide data has been investigated by Shanxi Geological Environment Monitoring Center; adopted the 5-fold cross validation method and through the Geostatistics analysis means the datasets of all non-landslide and landslide were divided into 80:20 proportions randomly for training and validating models. 15 condition factors including Terrain, Geological, Hydrological, land cover and human engineering activity factors (distance to road, Distance to mined area, Surface collapse density) are selected as the evaluation indices to construct susceptibility assessment model. Three Machine learning algorithms for landslide susceptibility prediction (LSP) including C5.0 Decision Tree (C5.0), Random Forest (RF) and Support Vector Machine (SVM) have been selected and compared through the Areas under the Receiver Operating Characteristics (ROC) Curves (AUC), and several statistical estimates. The study revealed that for these three models the value range of prediction accuracies vary from 83.49% to 99.29% (in training stage), and 62.26% to 73.58% (in validating stage). In the two stages, AUCs are between 0.92 to 0.99 and 0.71 to 0.80 respectively. Using Jenks Natural Breaks algorithm, three LSPs levels are established as very low, low, medium, high and very high probability of landslide by dividing the indices of LSP. Compared with RF and SVM, C5.0 is considered better in five categories according to quantities and distribution of the landslides and their area percentage for different LSP zones. Four factors such as distance to road, lithology, Profile curvature and Surface collapse density are the most suitable condition factors for LSP. The distance to mine area factor has medium contribution, and the surface collapse density factor plays an obvious role in the occurrence of geological hazards to all the models. The result reveals that C5.0 possesses better prediction efficiency than RF and SVM, and Underground mined area and surface collapse affect significantly on the occurrence of landslides in Fenxi Coal Mine Area.